Predicting Crime without the Pre-Cogs?

Crime in Chicago is difficult to predict. Hyde Park is remarkably safer than when I was a student here a decade ago (and in fact is one of the city's safest neighborhoods), but this past week saw a troubling and sad increase in violence in other parts of the city. The Mayor and police attributed some of the increase to the warm weather, but policy makers and citizens seemed surprised at the spike, which goes against national and regional trends. Predicting when crime will occur and the impact that different policing and punishment policies will have on crime is a tricky business. So how are policy makers supposed to decide how many resources to deploy, where to deploy them, and what the impact of, say, a moratorium on the death penalty (as we had here in Illinois recently) will have on crime?

Fiction offers us one possibility. In the movie version of Philip Dick's story "Minority Report," a police officer played by Tom Cruise leads a Division of Pre-Crime that with the use of three semi-conscious "pre-cogs" can predict certain crimes before they happen. When the pre-cogs "see" a murder a few minutes in the future, the police rush to arrest the perpetrator before the act is committed. The movie, which is fascinating and brilliantly done, raises a host of legal issues, not the least of which is whether punishment of mens rea without actus reus is sensible public policy. But we don't have pre-cogs. (At least not that I know of.)

In the absence of pre-cognitive superbeings and Tom Cruise, what are the alternatives? There is a vibrant academic literature on predicting crime, with models of various types offered as the best way of estimating future crime rates. Many of these involve mapping software, which plots the past in the hopes of extrapolating to the future. Police use some of these techniques, but most are very crude, using things like weather or the location of liquor stores as "hot spots" to estimate crime rates. Police, we are told by them, also use experience and gut instinct. All of the various methods, whether formal models or inside the head of the commissioner of police, are deployed in haphazard and isolated ways.

In a paper recently posted to SSRN, co-authors Justin Wolfers (Wharton) and Eric Zitzewitz (Dartmouth) and I offer an alternative that is designed to extract from the "market" estimates that aggregate and incorporate all readily available information about future crime rates under various policy choices. Here is the abstract:

Prediction markets have been proposed for a variety of public policy purposes, but no one has considered their application in perhaps the most obvious policy area: crime. This paper proposes and examines the use of prediction markets to forecast crime rates and the impact on crime from changes to crime policy, such as resource allocation, policing strategies, sentencing, post-conviction treatment, and so on. We make several contributions to the prediction markets and crime forecasting literature.

First, we argue that prediction markets are especially useful in crime rate forecasting and criminal policy analysis, because information relevant to decisionmakers is voluminous, dispersed, and difficult to process efficiently. After surveying the current forecasting practices and techniques, we examine the use of standard prediction markets - such as those being used to predict everything from the weather to political elections to flu outbreaks - as a method of forecasting crime rates of various kinds.

Second, we introduce some theoretical improvements to existing prediction markets that are designed to address specific issues that arise in policy-making applications, such as crime rate forecasting. Specifically, we develop the idea of prediction market event studies that can be used to test the influence of policy changes, both real and hypothetical, on crime rates. Given the high costs of changing policies, say issuing a moratorium on the death penalty or lowering mandatory minimum sentences for certain crimes, these markets provide a useful tool for policy makers operating under uncertainty.

These event studies and the other policy markets we propose face a big hurdle, however, because predictions about the future embed assumptions about the very policy choices they are designed to measure. We offer a method by which policy makers can interpret market forecasts in a way that isolates or unpacks underlying crime factors from expected policy responses, even when the responses are dependent on the crime factors.

Finally, we discuss some practical issues about designing these markets, such as how to ensure liquidity, how to structure contracts, and the optimal market scope. We conclude with a modest proposal for experimenting with markets in this policy area.

Comments

[Roach]

It is and it is not difficult to predict crime. Little spikes of crime as to locale and timing may be hard to predict in terms of timing. But neighborhoods where crime will occur is easy to predict.

Areas with high numbers of fatherless minority children and young men have astronomical crime rates. Blacks have a crime rate 7X as high as whites. Hispanics something like 3.5X. Poor, illegitimate blacks and Hispanics exceed even these numbers on a per capita basis. Areas filled with white (and Asian) married couples and their children have very low crime. Areas differing on one or another of these factors--age, rate of illegitimacy, and ethnicity--vary accordingly.

But crime generally speaking is easy to predict, housing values reflect this fact, and so do our intuitions. It's not for nothing that Obama's Reverend lives in a 10,000 square foot mansion in a white part of town.

There's no doubt cops informally practice racial profilings, stopping blacks in white areas and whites in black areas. But a deliberate, demographic-driven allocation of police resources might make sense. For instance, black clubs and black concerts and things like the OJ verdict demand more resources than country western bars, Van Halen concerts, and the Scooter Libby Verdict.

Yes, yes, I know this is all terribly insensitive. It happens to be true, though, and the statistics aren't hard to come by.

[Todd Henderson]

Roach:

You are talking about baselines, and your points may be accurate. I don't know. Assuming they are, this still doesn't tell police or policy makers what the changes (plus or minus) from baselines will be next week or next month or next year. Surely crime rates are not steady over any time period, and it is this variation that individuals who allocate resources, be they patrol cars or prosecutors, want to understand.

In addition, the paper addresses how to consider issues like changes in criminal policy, such as sentencing reforms or the death penalty. A governor would want to know the expected increase in the crime rate if she issues a moratorium on the death penalty, but the existing techniques (e.g., retrospective academic studies, gut instinct, experience from other countries) are likely to be suboptimal considered alone, if considered at all. Markets are the best aggregators of information, whatever its topic or source. This is especially true, as here, where there are significant barriers to the efficient sharing and processing of information.

[Roach]

A few quick thoughts.

Do we want police to be reallocated on a daily or weekly basis? It seems to me part of the sensible rationale behind community policing is that police in the same area knowing the same people develop a sense of ownership over their "beat" and also develop informal informational networks. They should clearly be allocated so that arrest rates/crime rates are in rough proportion to one another. It's unreasonable to expect uniform crime rates, but you wouldn't want the arrest/crime ration to get way out of wack.

Second, any "markets" or preditive software or police intelligence apparatus is likely to be bogged down with various politically correct "no go zones," most especially involving race. Imagine how much more private industry insurance risk calculations would be if race, and not mere proxies like zip code, could be used. This is the Elephant in the room, and it's a shame that academics--even tenured ones like James Watson--are essentially unable to speak about this in our "Free Country." We're as skittish today about race as Victorians were about sex.

Three, I'm skeptical of the claim that markets are always the most efficient means of aggregating data. Markets are efficient at pricing things that people pay for. But people have other ways of receiving and sharing information in non-market areas--like war, dating, crime, etc.--most especially the gift of intuition. I realize the goal of social science is to move beyond mere intuition to actual knowledge, but consider all the ways we know a neighborhood is "bad": quality of real estate, clothing, number of aimless young men, quality of cars, loudness of music, signs of disorder, number of police, number of young children with unescorted mothers, Newport t-shirts, boarded up businesses, etc. In other words, it seems most people (and therefore most cops) have a pretty good sense of where their attention is needed or where they should avoid contact in the case of a citizen.

It's likely that these intuitions which translate into behavior are as valuable as any derivative market like a prediction market. (I'm thinking particularly of the insights of someone like Gavin DeBecker in his book "The Gift of Fear.") In other words, I think that derivative markets--in this case, people making predictions with no stake in the outcome--are likely to be more uncertain and more error-prone that the kind of implicit economics involved in revealed preferences, for example, the use of the club anti-theft device, putting bars on their windows, or buying a gun.

This is to say the predictive markets, like the stock markets, are only as good as their participants. Notice the lemming-like behavior in mortgage markets that have preceded our current disaster. The key to a predictive market's function are the players' underlying models and methodologies. These models and methods tapping into existing though inchoate knowledge could be easily employed and used by cops and policy markers. They already are in the form of informal folk knowledge, intuition, tips, and the like.

I used to have a bumper sticker that said: "Fight Crime, Shoot Back." You don't need regression analysis to know the crime-fighting impact of that on any would-be car jackers.

[Joan A. Conway, Pro Se Litigant In Forma Pauperis]

Could the Dirksen Building and DePaul's loop campus bomb and package scares today, April 21, 2008, have anything to do with the tight election and the primary in Pennsylvania tomorrow?

Many people have predicted that the fight between Hillary and Obama will escalate and even if it is Obama in a competition with McCain in November the treats to our safety will not abate, but increase and retribution will visit upon us after the election, either way.

Since a victory for Obama, or his failure to win the Presidency can ignite hostilities, especially in this city in celebration or in retaliation.

"But crime generally speaking is easy to predict, housing values reflect this fact, and so do our intuitions. It's not for nothing that Obama's Reverend lives in a 10,000 square foot mansion in a white part of town" stated Roach.

There is a bond between business owners and landlords in wealthy neighborhoods that can bribe the local police captain and his force, or the fire captain, and his men, to see that they are given extra care and security, even when the use of protocol obscures the fact that the business owners and/or landlords caused the disturbance to frighten a certain element threatening their livelihood, such as a local bar, or set building and hotel fire(s) to intimate tenants when there is labor unrest.

Does any market model make the distinction that authorities are often corrupt and allow crime to go undetected which can be associated with management?

Not all crime, especially in a city as diverse as Chicago, stems from its victims.

And often the authority is in cahouts with the perpetrators of these instances of crime.

I have seen this with my own eyes how the fire department questions some of the inhabitants of slum hotels while overlooking some of the other inhabitants making management accusations of a breach of duty, and loyalty, who started the fire in the first place.

[Sue Hough]

Your insight into the Chicago "crime scene" is wonderful. I am hoping you will be able to direct me to the proper source of information for my report. I am preparing an evaluation of current crime rates within a small neighborhood in Chicago, the 8000 block of South Marshfield and the surrounding 6 blocks. I need to create comparisons over the past 5 years. So far, I have been unable to narrow my search to this level. Can you direct me to the proper sources?